Getting started with Houseware's metric packages

We assume that you have an understanding of dbt-core before you dive into dbt-metrics. If not, check out this amazing documentation dbt folks have written.

Why?

With an abundance of data, there is an increasing inability to make sense of it. From data pipelining to data cleaning and classification, converting raw data into business readable tables is a full-fledged data engineering job requiring sweat and tears.

We realized that the end-state being metrics, is all that a user cares about to make data-driven decisions, and the process for business users to obtain these metrics is ugly and sweaty and needs to be simplified and abstracted away. Enter dbt-metric packages.

Metrics! Precise, flexible, and accessible with consistency and precision in the tools you know and love on day one.

Drew Banin, July 2022 | Read more about the semantic layer

dbt-metrics

dbt metrics ensure metric consistency and provide a way to standardize metrics under version control in dbt projects. By abstracting metrics calculations out of pre-aggregated tables or specific business intelligence tools (BI tools), dbt metrics can be defined once and used everywhere. Defining metrics in one place ensures consistent reporting of key business metrics, especially in an environment where metric definitions and dimensions are changing along with your business

Check-out dbt's documentation on metrics and how to use them!

About Houseware's metric packages

Our metric packages work for any organization which uses Fivetran to ETL data from SaaS tools. Data from SaaS tools are raw and require a lot of data engineering for business users to make sense of it. We eliminate this friction by pre-computing standard metrics out of these sources. These metrics are stored inside your warehouse inside the schema/dataset you provide during your dbt project setup.

How to get these metrics?

The metrics reside in your data warehouse and can be directly used in any SaaS tool you use for your data needs. You can plug them into any BI tool or R-ETL them into your business software.

To get started, you first need to choose the SaaS tool you want to do the metric transformations on. You can select from the ones available at metrics.houseware.io.

Importing packages

You then create a packages.yml file for your dbt project, including the metric package.(You can add multiple houseware metric packages)

packages:

- git: "https://github.com/HousewareHQ/dbt_segment_metrics.git"
 revision: v1.0.1packages.yml

Running the package

You will need to run the dbt deps to install the packages. Post that, a dbt run will trigger all the models and metrics to be computed. The transformations will take place in the target schema and target database you provided while configuring the dbt project.

Blazing fast metrics in your warehouse ⚡

That's all! You can find your metrics sitting on top of your warehouse! You can now share these metric tables with your revenue teams in the fastest way possible.

Why we open-sourced our metric packages and what we have ahead

We believe in metrics being a first-class citizen for business teams and understand how metrics are firstly challenging to obtain and not consistent amongst groups. We wanted to solve this problem for everyone. We also wanted to open contributions from the vibrant dbt community so that all of us can work together on making metric-driven data experiences accessible to all.

What next?

We plan to release as many packages to get out-of-the-box metrics for all SaaS tools that business teams use. With our shipping velocity, having more members from the community help us with additional metrics, QA, and packages would help us accelerate our vision of data accessibility.

Heavily inspired by what dbt is doing for the data community and how Fivetran builds open-source dbt packages, this is our way of contributing to the data community.

How do I contribute?

1. Go to metrics.houseware.io
2. Find a package you're interested in contributing to, go to the corresponding GitHub repository, and explore the metric box.
3. Go through the metric package and identify issues you want to contribute to. These can be already existing issues or new improvements you find and want to work on.